PII S0145-2134(98)00056-8

PRACTITIONERS WHO INTERVIEW children often ask for an accessible introduction to the research ndings on
suggestibility or the accuracy of childrens testimony, as well as a summary of the developmental aspects of their task. The
following two annotations have been commissioned to ll that gap, especially with non-psychologists in mind. Many
investigative interviews are undertaken by police ofcers or social workers, with only a minority by child psychologists or
psychiatrists. In addition, research in this area has moved very fast, and for this reason too, the two reviews are likely to
prove helpful to practitioners. Lamb and colleagues review the available research on childrens ability to relay their
experiences to interviewers, and what can be done to enhance that capacity. Saywitz summarizes research from develop-
mental psychology, pertinent to the interview in these situations. We hope that these two articles, taken together, will
provide a helpful basis for the difcult yet crucial task of investigative interviewing with children who have witnessed or
experienced traumatic events.
SPOTLIGHT ON PRACTICEANNOTATION
CONDUCTING INVESTIGATIVE INTERVIEWS OF
ALLEGED SEXUAL ABUSE VICTIMS
MICHAEL E. LAMB AND KATHLEEN J. STERNBERG
Section on Social and Emotional Development, National Institute of Child Health and Human Development, Bethesda,
MD, USA
PHILLIP W. ESPLIN
Private Practice, Phoenix, AZ, USA
ABSTRACT
Objectives: There were two aims: First, to describe the factors that inuence childrens competence and second, to discuss
ways in which investigative interviewers can maximize the quality and quantity of information they obtain from alleged
witnesses and victims.
Method: No new research is described in this paper. Rather, the authors provide a focused review of the relevant literature
designed to be maximally useful for practitioners.
Conclusions: Children are often the only available sources of information about possible abusive experiences. Research has
shown that children can, in fact, be remarkably competent informants, although the quality and quantity of the information
they provide is greatly inuenced by the ways in which they are interviewed. This article describes ways in which
investigative interviewers can maximize the amount and quality of information they elicit from alleged victims. 1998
Elsevier Science Ltd
Received for publication November 4, 1997; nal revision received November 5, 1997; accepted November 5, 1997.
Reprint requests should be addressed to Michael E. Lamb, Section on Social and Emotional Development, National Institute
of Child Health and Human Development, 9190 Rockville Pike, Bethesda, MD 20814.
Pergamon
Child Abuse & Neglect, Vol. 22, No. 8, pp. 813823, 1998
Copyright 1998 Elsevier Science Ltd
Printed in the USA. All rights reserved
0145-2134/98 $19.00 .00
813
INTRODUCTION
SEX CRIMES AGAINST children are not a new phenomenon, though there have been dramatic
increases in the number of alleged victims over the last several years. The increased awareness of
sexual abuse has in turn fostered extensive efforts to improve the quality of forensic investigation,
with researchers and practitioners eager to identify reliable means of determining whether or not
children have been abused. Many experts once believed that medical or physical evidence would
prove denitive, though these hopes have not been borne out. Indeed, the misidentication of
victims on the basis of physical symptoms, such as anal dilatation, now known to be unreliable
indicators of prior abuse has fostered systematic reevaluation of forensic and social welfare
practices (e.g., Butler-Sloss, 1988) and has led in turn to a focus on children as sources of
information about their experiences (Lamb, 1994). Sex crimes against children are extremely
difcult to investigate precisely because the only evidence often consists of the victims and
suspects accounts of the alleged events. Alleged perpetrators are likely to deny or misrepresent
their behavior, and thus children become crucial sources of information. Unfortunately, childrens
statements are not always easy to elicit or interpret and can often be misinterpreted or misused. The
incidents Ceci and Bruck (1995) discuss, along with recent increases in the numbers of children
who allege that they have been victims of sexual abuse, underscore the need for reliable techniques
to enhance the quality and quantity of information obtained from young alleged victims and
witnesses. As many frustrated interviewers, lawyers, and judges can attest, however, childrens
accounts of their experiences are frequently quite skeletal and even contradictory, and this can raise
doubts about the childrens competence, especially when the offenses are serious and the possible
consequences for alleged perpetrators are severe. Not surprisingly, this realization often prompts
questions about the alleged victims competence and credibility, and as a result considerable
research has been conducted on the memorial, communicative, and social tendencies and abilities
that inuence childrens competence as informants.
As Lamb, Sternberg, and Esplin (1994) noted in their review, ve central factors profoundly
affect childrens capacities as witnesses. First, children tend to be be reticent with unfamiliar adults,
and this may make them uncommunicative. Second, children are used to being tested by adults
(What color are my shoes? Do you remember what we were going to do today?) but are seldom
treated by adults as unique sources of otherwise unavailable information. Both of these charac-
teristics make it necessary to motivate potential witnesses to be as informative and detailed as
possible. Third, children have poorer linguistic skills than adults; they may use words idiosyncrat-
ically, their vocabularies are more restricted, and their sentences tend to be abbreviated, shorn of
extensive elaboration. Fourth, their memories are not as good as those of adults, with the amount
of information remembered gradually increasing with age. In addition, probably because they have
had fewer experiences with which to associate new information to make it more memorable,
children tend to forget more rapidly than adults do, and this makes it especially important to
conduct investigative interviews as soon as possible after the alleged events have taken place. On
the other hand, the memories of children are not more prone to error than those of adults, so that,
although they may remember less, they are about as accurate, making roughly the same proportion
of errors as adults or older children do. As with adults, furthermore, information retrieved from
recognition memory is more prone to error than freely recalled information, and it is thus important
to retrieve as much information as possible from recall memory when conducting forensic
interviews. Fifth, although both children and adults are suggestible, preschoolers appear especially
susceptible, particularly to post-event contamination. Suggestions are less likely to affect childrens
responses when they pertain to central or salient details of salient events or to appearances rather
than the sequence of events. By contrast, suggestions may be more inuential when the memory
is not rich or recent, when the questions themselves are so complicated that the witness is confused,
and when the interviewer appears to have such authority or status that the witness feels compelled
814 M. E. Lamb, K. J. Sternberg, and P. W. Esplin
to accept his or her implied construction of the events. We must also avoid holding children to a
higher standard than adult witnesses, while simultaneously questioning their competence and
credibility. Adults, like children, respond to coercion, suggestion, and manipulation, and children
are not well served by the implicit assumption that they hold within their minds more information
than they are able to provide. It is especially important to interview young children as soon as
possible after the alleged or suspected events (the passage of time may affect both memory and the
susceptibility to suggestion) and to maintain an electronic record of the interview to help refute post
hoc allegations that the childs account contains details wittingly or unwittingly suggested by the
interviewer. Readers are referred to Ceci and Bruck (1993, 1995), Lamb and colleagues (1994),
Perry and Wrightsman (1991), and Poole and Lamb (1998) for reviews of the evidence supporting
the statements made in this paragraph.
Fantasy, memory strategies and deciencies, suggestibility, and communicative abilities impor-
tantly affect the accounts provided by young children of their experiences but children can and do
remember important details of incidents that they have observed or experienced. Although their
accounts can be manipulated, sensitive interviewers who are aware of childrens capacities and
deciencies can avoid many of the problems posed by questions that force children to operate at
or beyond the limits of their capacities. Linguistic and memorial difculties do not make children
incompetent witnesses, but an understanding of their capacities and limitations should inuence the
ways in which children are interviewed and the ways in which their accounts are interpreted.
Likewise, the demonstrable fact that investigative interviews with young children can be rendered
worthless by inept practice should not blind us to the substantial literature demonstrating that
reliable information can be elicited from young children who are competently interviewed. Our
goal in this brief article is to articulate strategies for eliciting as much reliable information as
possible from young children. This emphasis reects our rm belief that the informativeness of
interviews with child victims is strongly inuenced by the skill and expertise of the interviewers
and that skillful interviewers can help make children into reliable and invaluable informants.
A variety of professional groups and researchers have offered recommendations regarding the
most effective ways of conducting forensic or investigative interviews (American Professional
Society on the Abuse of Children (APSAC), 1990; Fisher & Geiselman, 1992; Jones, 1992; Lamb
et al., 1994, 1995; Memorandum of Good Practice, 1992; Raskin & Esplin, 1991a; Raskin &
Yuille, 1989). As Poole and Lamb (1998) pointed out, these statements reveal a substantial degree
of consensus regarding the ways in which investigative interviews should be conducted, and the
recommendations offered in this paper reect this remarkable consensus. We suggest ways in
which forensic interviewers can obtain the greatest amount of reliable information from young
children, offering recommendations that are informed by research conducted in both laboratory
analog and eld settings, even though for ease of exposition we refer to few of the relevant studies
directly.
Research on Investigative Interviews
Unfortunately, agreement about the goals and desired characteristics of investigative interviews
have not ensured that forensic interviews are done uniformly well. In eld research on front-line
investigative interviews, Lamb, Hershkowitz, Sternberg, Esplin, Hovav, Manor, and Yudilevitch
(1996) and Sternberg, Lamb, Hershkowitz, Esplin, Redlich, and Sunshine (1996) distinguished
among 10 types of interviewer utterances, including three that involved focusing the childs
attention, sometimes in a way that suggested a desired response, and two that involved nondirective
prompts (open-ended invitations and facilitators). The three types of focused utterances lay along
a continuum of risk, in that they varied with respect to the degree of suggestive inuence they
exerted on childrens responses while tapping the childs recognition memory. We found that, in
investigative interviews, invitations elicited responses from children that were, on average, three
Conducting investigative interviews 815
times longer and three times more detailed than did any of the more focused prompts; the
superiority of open-ended utterances was apparent regardless of the age of the children being
interviewed. Because focused questions draw information from recognition rather than recall
memory, furthermore, it is more likely to be inaccurate, although the value can be maximized and
the risks minimized if each focused question (tapping recognition memory) is followed by (paired
with) an open-ended question placing the burden back on recall memory. Unfortunately, focused
utterances are much more common in the eld than open-ended questions are. In the eld sites
studied, more than 80% of the interviewer utterances were focused whereas only 6% were
invitations. Research in the United States, the United Kingdom, and Israel shows that the
over-reliance on focused questions is evident regardless of the childrens age, the nature of the
offenses, the professional background of the interviewers, or the utilization of props and tools like
anatomical dolls (Craig, Sheibe, Kircher, Raskin, & Dodd, 1996; Lamb, Hershkowitz, Sternberg,
Boat, & Everson, 1996; Lamb, Hershkowitz, Sternberg, Esplin, Hovav, Manor, & Yudilevitch,
1996; Sternberg, Lamb, Hershkowitz, Esplin, Redlich, & Sunshine, 1996; Walker & Hunt, in press;
Westcott, Davies, & Horan, 1998). Overall, narrative responses are more desirable because they are
more detailed, because they are more accurate, and because they are necessary for systematic
evaluation of credibility using techniques like the Criterion-Based Content Analysis procedure
(Hershkowitz, Lamb, Sternberg, & Esplin, 1997; Lamb, Sternberg, Esplin, Hershkowitz, Orbach,
& Hovav, 1997; Raskin & Esplin, 1991a, 1991b).
Changing interview practices is quite difcult, however, as our own experiences illustrate quite
well. Recognizing that forensic interviewers tended to be too reliant on focused questions, we urged
interviewers attending workshops we conducted to use more open-ended questions in their
interviews, particularly when beginning the substantive phase of the interview when narrative
accounts of the alleged abuse are very important. In intensive training seminars (approximately 40
hours in length) for groups of investigators who had agreed to participate in our studies, we
described memory processes, reviewed childrens linguistic and memory capacities, discussed
factors inuencing suggestibility, and suggested that interviews be organized so as to include
rapport-building phases, substantive phases, and closure phases in succession. We also reviewed
videotapes of forensic interviews that illustrated the appropriate and inappropriate use of both
open-ended and focused questions. Participants were encouraged to ask questions and a great deal
of time was devoted to discussions. In addition, we described the conceptual bases of Statement
Validity Analysis and Criterion-Based Content Analysis as we expected that familiarity with these
techniques might improve interview quality, as suggested by Undeutsch (1982, 1989) and Raskin
and Yuille (1989). Written feedback was later provided on transcripts of interviews conducted after
the training.
In spite of our consistent emphasis on the importance of obtaining as much information as
possible from free recall memory, the interviewers who participated in our rst few studies
continued to rely on focused questions to elicit information from children (Lamb, Hershkowitz,
Sternberg, Esplin, Hovav, Manor, & Yudilevitch, 1996; Sternberg, Lamb, Hershkowitz, Esplin,
Redlich, & Sunshine, 1996). These unexpected ndings underscored how difcult it was for many
of these interviewers to obtain information from children. Each interview is unique and involves
a tough juggling act on the part of an interviewer who is trying to determine what may have
happened to the child while attempting to ask nonleading questions. The challenges are exacerbated
when interviewers have had little experience working with children, receive little or no formal
training, conduct investigative interviews of alleged victims infrequently, and rarely review their
interviews.
In light of these difculties, we designed a study which required forensic interviewers to follow
very specic scripts in the rapport-building phases of their interviews (Sternberg, Lamb,
Hershkowitz, Yudilevitch, Orbach, Esplin, & Hovav, 1997). The goal of this study was to evaluate
the relative effectiveness, in forensic rather than analog contexts, of two techniques for motivating
816 M. E. Lamb, K. J. Sternberg, and P. W. Esplin
young witnesses to provide detailed accounts of alleged experiences of sexual abuse. In half of the
interviews, investigators who were naive with respect to the experimental hypotheses used a script
containing many open-ended utterances to establish rapport, whereas in the other interviews the
same investigators used a script involving many direct questions. Both introductory scripts took
about 7 minutes to complete and both included the identical open-ended statement to initiate the
substantive phase of the interview. When the resulting interviews were examined, we found that
children who had been trained in the open-ended condition provided two and one-half times as
many details and words in response to the rst substantive utterance as did children in the direct
introduction condition. Children in the open-ended condition continued to provide more informa-
tion in response to subsequent invitations, suggesting that the initial training was successful in
conveying the interviewers desire for detailed description of the alleged events. Evidently,
children who had the opportunity to practice providing lengthy narrative responses to open-ended
questions in the introductory phase of the interview continued this pattern after their interviewers
shifted focus to the alleged incidents of abuse.
We were quite impressed that the richness of childrens accounts could be inuenced by the
interrogatory style modeled in brief introductory segments of the interview, and that a single
carefully worded prompt could elicit so much information from the children in both conditions. In
Sternberg and colleagues (1997) study, the rst substantive question yielded an average of 38
details from children in the direct introduction condition and 91 details from children in the
open-ended introduction conditions. In an earlier nonexperimental study involving interviews by
the same group of investigators interviewing children of the same age who had experienced similar
types of abuse, the average invitation yielded only ve details, and the rst invitation in that study
yielded an average of only six details (Lamb, Hershkowitz, Sternberg, Esplin, Hovav, Manor, &
Yudilevitch, 1996). The discrepancy between these averages underscores the value of detailed
interview protocols, and of encouraging interviewers to use prompts that are as open-ended and
nonsuggestive as possible.
Interestingly, although the open-ended training inuenced the response style of the children who
participated in this study, it had little effect on the interviewers style of questioning after the rst
substantive question was posed. In other words, even when children provided lengthy responses to
the rst open-ended substantive question, interviewers did not continue to ask open-ended ques-
tions but rather shifted to more focused questions. This unexpected nding suggested that it might
be valuable to script additional open-ended questions throughout the substantive phase of the
interview and thus we developed increasingly detailed scripts for the entire interview (including
substantive and non-substantive sections) that are currently being tested in the eld.
Preliminary ndings suggest that these extended scripts indeed improve the overall informa-
tiveness of forensic interviews. Interviewers retrieve more information using open-ended questions,
conduct better organized interviews, and are more likely to follow focused questions with
open-ended probes (pairing), as we suggested. Interviewers clearly have difculty internalizing
recommended interview techniques and may need more explicit guidelines than those typically
provided in training sessions or manuals, however intensive. In eld settings, interviewers who
follow scripts seem to elicit more information from recall memory and to avoid more potentially
dangerous interviewing practices than do interviewers who improvise, despite the apparent disad-
vantages of inexible standardized scripts. In addition, we cannot overemphasize the value of
continued peer review, training, and the systematic analysis of videotaped and transcribed inter-
views.
Evaluation of the information obtained in an investigative interview can only proceed when there
is a complete electronic record preferably a videotape not only of the childs responses but
also of the prompts by which they were elicited and the relative location of details derived from
recall memory and those details elicited by more suggestive prompts (Lamb, 1994). (A suggestive
utterance 20 minutes into the interview clearly does not reduce the value of a narrative description
Conducting investigative interviews 817
provided at the beginning of the interview, for example.) Although there has been considerable
professional fear that electronic records permit defense attorneys to unfairly impugn the value of
childrens testimony, the electronic record of a competent interview is clearly of greater value to
prosecutors than to defense attorneys because it permits illustration of the entire interview process
and prevents the selective and unfair focus on single responses or suggestive questions taken out
of context. Furthermore, electronic records of interviews conducted competently shortly after the
alleged events provide a permanent record of information obtained from a witness whose ability to
remember the events is certain to decline over time, and whose account may easily be affected by
post-event contamination in the course of repetitive interviewing by professionals and family
members (Ornstein, Larus, & Clubb, 1991). Although electronic records cannot replace in vivo
testimony at the time of trial, they can prove to be remarkably damning pieces of evidence in the
eyes of juries, judges, and even defendants.
The ndings reported by Sternberg and colleagues (1997) are consistent with the results of
laboratory/analog studies suggesting that motivational and contextual factors play an important role
in shaping childrens reports of experienced events (Paris, 1988; Saywitz, Goodman, Nicholas, &
Moan, 1991). Along with the results of our ongoing research, they also suggest that, even in
authentic forensic interviews, it is possible to entrain response styles that enhance the richness of
information provided by children by ensuring them the opportunity to practice a narrative style and
by reinforcing this style in the pre-substantive portion of the interviews. Richly detailed accounts
of abusive events facilitate the more effective investigation of crimes and provide child protection
workers with more information upon which to base their evaluations. The results of this study also
highlight the value of having interviewers clearly communicate their expectations concerning the
childs role. On the other hand, the studies we have conducted thus far have several important
limitations. First, although open-ended interview procedures elicit more information from children
of all ages studied, we have not been able to evaluate the accuracy of the information provided. It
would thus be valuable to compare the direct and open-ended interview scripts in an analog study
in which children described known events. Second, we do not yet know whether our ndings can
be generalized to children under 5 years of age. Although there is consensus among researchers and
clinicians that children under 5 are the most difcult to interview, systematic eld research on
preschoolers in forensic contexts is scarce.
As noted above, several studies are currently being conducted to explore further the utility of
open-ended questions in the substantive phase of the interview. These studies should help
determine when it might be necessary to use more focused utterances, perhaps paired with
follow-up open-ended probes as suggested by Lamb and colleagues (1995) and Jones (1992). In the
interim, we offer below some suggestions regarding the structure, organization, and content of
investigative interviews based on the results of relevant laboratory/analog, developmental, and eld
studies.
Conducting Investigative Interviews
The purpose of investigative interviews is to explore and evaluate alternative hypotheses, and
interviewers should thus prepare themselves by gathering as much information as possible about
the alleged incidents, the interviewees capacities and propensities, and their motivations to be
honest, deceptive, or misleading (Green, 1991; Myers, 1994; Perry & Wrightsman, 1991; Raskin
& Yuille, 1989). Careful preparation enables interviewers to rene and evaluate their hypotheses
as their interviews progress.
Because preschoolers, and even children in the early elementary grades, use language more
idiosyncratically and less maturely than older children, it is also crucial to evaluate childrens
linguistic styles and skills informally before interviews with them begin. Interviewers may want to
pay attention to informal conversations between children and familiar adults accompanying them
818 M. E. Lamb, K. J. Sternberg, and P. W. Esplin
in order to gauge linguistic competence. This information can later help interviewers: (a) determine
whether and when rapport has been established; (b) frame questions using developmentally and
individually appropriate language; (c) evaluate uency when describing neutral topics (Raskin &
Yuille, 1989; Saywitz, 1987; Walker, 1993); and (d) protect interviewers from becoming exasper-
ated by the brevity of childrens responses and then being tempted to ask too many focused
questions, which not only reduce the amount of information obtained, but may also lead witnesses
to impeach themselves.
Because they are also unused to being treated as informants and may be reticent with or afraid
of strangers and investigators, young children need to be motivated much more carefully and
extensively than older children and adults (Saywitz, 1987). Not only does their performance depend
on feeling comfortable enough with interviewers to describe intimate and possibly embarrassing
events, it also depends on their recognition that these adults really do value what they have to say.
In most everyday contexts, adults test children by asking questions to which they already know
the answers, whereas in forensic interviews children are potential sources of novel information.
Furthermore, because children often feel obliged to agree with or at least respond to adults
questions or assertions (Hughes & Grieve, 1980), some investigators recommend explaining to
children in the pre-substantive phase that answers like I dont know or I cant remember are
acceptable, and that children should correct interviewers who appear to misunderstand them
(Memorandum of Good Practice, 1992). All experts agree that interviewers must avoid pressuring
children to acquiesce and encourage them to dispute false suggestions.
To establish rapport while simultaneously fostering a response set in which descriptions from
recall memory rather than yes/no responses predominate, we recommend that children be asked to
recall and describe some recent meaningful event, such as their last birthday party or a memorable
holiday, in detail. Reviewing that party or holiday and encouraging children to really tell
everything about it graphically illustrate that the interviewers expect to hear detailed narrative
accounts and are interested in the childrens experiences. If children provide very brief accounts or
only describe part of the birthday party, interviewers can encourage children to provide more
information, thereby training them to provide detailed accounts (e.g., of the sexual allegation),
while informally evaluating their linguistic, expressive, and descriptive capacities. As explained
above, our research conrms that such training leads to increases in the amounts of information
provided by alleged victims in response to the rst substantive utterance by interviewers (Sternberg
et al., 1997).
After completing the rapport-building phase of the interview, investigators need to guide
children in a nonsuggestive fashion to focus on the allegations of sexual abuse. This is not always
easy, especially with young children, and may require extensive creativity on the part of inter-
viewers, who must explain the scope of information needed, rather than assume that the children
know what information is relevant. The transition from the rapport-building to the substantive
portions of the interview can often be accomplished by saying something like: I understand that
something may have happened to you yesterday. Please tell me about that. Unfortunately,
however, childrens responses are often very brief and interviewers must thus probe further to
signal that they are interested in detailed descriptions of specic incidents. Interviewers can pursue
further information from recall memory by referring to a salient component of the childs account
(e.g., Earlier you said something about a bed. Tell me everything about that. or You mentioned
that it happened at Grandmas house. Tell me everything that happened from the minute that you
got to Grandmas.) Investigators can also elicit more information by feigning confusion (Gee,
Im confused. You said that Mr. B. was in your bed. Tell me how that happened.) By asking
children for more information and by feigning confusion, interviewers can empower children to be
better informants by conveying interest in complete, detailed accounts.
Although open-ended questions are most likely to encourage accurate accounts of events
children have experienced, these accounts are often incomplete, especially when preschoolers are
Conducting investigative interviews 819
being interviewed. As a result, it is often necessary to begin asking focused questions quite early
in the interviews of young children (e.g., Lamb et al., 1994; Myers, 1994; Saywitz, 1987; Spencer
& Flin, 1990). Whenever it is necessary to use questions that focus the childs attention on certain
events, people, or places, however, they should be followed by (paired with) open-ended
questions designed to elicit free narratives about the topic to which the interviewer wishes to direct
the childs attention (Jones, 1992; Lamb et al., 1994, 1995; Raskin & Esplin, 1991a). For example,
Did anything ever happen in the living room? can be followed by Tell me everything that
happened there. This strategy reduces reliance on recognition memory, which is more prone to
errors of commission, and emphasizes recall memory where the more plentiful errors of omission
are less likely to impeach witness credibility (Memorandum of Good Practice, 1992). Investigators
who employ such techniques, therefore, are less likely to misinterpret ambiguous or acquiescent
responses to their own statements.
For both evidentiary and protective purposes, it is often necessary to determine whether children
have been abused on one or multiple occasions, yet childrens responses may often be unclear in
this regard. Although children may be capable of describing specic incidents of abuse, whether
or not they do so may depend on the types of questions interviewers ask. The results of our recent
eld study suggest that the types of questions posed by interviewers partially determine whether
children provide general descriptions of abusive experiences or accounts of specic incidents
(Sternberg et al., 1996). Children know that adults typically expect relatively brief responses to
their questions, and thus respond to prompts such as Tell me whats been happening? with
summaries or scripts. Interviewers might determine whether multiple incidents occurred by asking:
Did this happen one time or more than one time? Young children nd it relatively easy to answer
this question, whereas questions like How many times, often lead children to impeach them-
selves. When children respond more than one time, interviewers can probe for additional
information by using time or location cues, such as references to the rst time, the last time,
or the time it happened in the barn. Unless they are probed in this manner, children may not go
beyond a general description of their abusive experiences. The interview scripts we have developed
and are currently rening in eld studies are designed to assist interviewers in obtaining the greatest
amount of information from young children, whether about a single experience or several such
experiences.
Although we clearly believe that open-ended prompts should be employed more frequently than
they typically are, it is important to recognize that even the most skillful investigators use direct and
leading questions when interviewing young children and that the inclusion of such questions does
not invalidate the testimony, provided that steps are taken to limit potential damage by framing
focused questions carefully, avoiding coercive repetition, or by pairing direct or leading questions
with open-ended prompts so as to return the child to recall (rather than recognition) memory.
The status of suggestive utterances is even more problematic, but not dissimilar. Children, like
adults, are clearly susceptible to suggestion, and at least in experimental contexts, preschoolers
appear especially susceptible (Ceci & Bruck, 1993), particularly to a form of suggestion
post-event contamination which involves the incorporation into later reports of incorrect details
suggested to them between the time of the incident and the time of the interview (Ceci, Ross, &
Toglia, 1987a, 1987b). In the face of repeated suggestion and coercion, it would not be surprising
if children incorporated erroneous information into their accounts, although this should not blind
us to the facts that even 3- to 5- year-olds are often resistant to noncoercive suggestion (Goodman
& Aman, 1990; Goodman, Aman, & Hirschman, 1987; Goodman, Bottoms, Schwartz-Kenney, &
Rudy, 1991; Goodman, Rudy, Bottoms, & Aman, 1990; Goodman, Wilson, Hazan, & Reed, 1989)
and that responses to single suggestive questions do not necessarily render childrens entire
statements dubious or invalid, just as adults statements are not invalidated by their suggestible
responses to some utterances. Suggestive utterances should be avoided whenever possible. When
a child fails to address certain issues in response to open-ended and directive prompts, however, it
820 M. E. Lamb, K. J. Sternberg, and P. W. Esplin
may be necessary for investigators to ask leading questions. For example, the police may have
obtained persuasive evidence (such as photographs, confessions, or medical examination results)
suggesting vaginal penetration that the alleged victim fails to mention. In such cases, a leading
question may be necessary. The potential damage attributable to a leading or suggestive question
will be minimized if: (a) the interviewer waits until the end of the interview (i.e., until she/he has
obtained narrative accounts of the incident and after exhausting less-damaging strategies); (b) the
suggestion is as limited as possible (Did anything ever happen to your vagina? is better than Did
he do anything to your vagina? which is, in turn, better than Did he ever put anything in your
vagina? and so on); and (c) a positive response to the suggestive or leading prompt is immediately
followed by an attempt to elicit further information from recall memory by pairing (Jones, 1992;
Lamb et al., 1994; Spencer & Flin, 1990).
CONCLUSION
We have attempted in this brief article to review our knowledge of the factors that inuence
childrens ability to provide reliable account of past experiences, and to show how interviewers can
take advantage of childrens strength, tendencies, and limitations when conducting investigative
interviews. There is a gradual increase with age in the likelihood that young children can be found
competent to testify, however. Few developmentalists would wonder, for example, whether a 1
year old could be competent or assume that (absent other disabilities) a 12 year old is likely to be
incompetent. Disagreement might well emerge, however, concerning the age at which most
children could be presumed competent. Most children of or under the age of 3 lack the commu-
nicative and memorial capacity to be competent witnesses, although competence rapidly increases
over the remaining preschool years and Jones and Krugman (1986) have persuasively documented
the ability of one 3 year old victim to provide compelling information about her experiences and
to identify her abuser. The rate of change in competence appears to decrease substantially once
children attain 6 years of age, with the majority of children in this age group being capable of
providing useful information when questioned competently (see Lamb et al., 1994 for a review).
Indeed, we have attempted to show in this article that children are often more competent informants
than adults realize. Interviews conducted in ways that clearly communicate the unique purposes of
investigative interviews, minimize the burden placed on childrens capacities, and take maximal
advantage of childrens abilities and our growing knowledge of memory and communication
demonstrate that children can be invaluable sources of information. Because alternative sources of
information about alleged or suspected events are seldom available, only improvement in the
average quality of investigative interviews are likely to bring about improvements in our ability to
protect children.
REFERENCES
American Professional Society on the Abuse of Children (1990). Guidelines for psychosocial evaluation of suspected sexual
abuse in young children. Chicago, IL: Author.
Butler-Sloss, E. (1988). Report of the inquiry into child abuse in Cleveland 1987. London, England: Her Majestys
Stationery Ofce.
Ceci, S. J., & Bruck, M. (1993). Suggestibility of the child witness: A historical review and synthesis. Psychological
Bulletin, 113, 403439.
Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A scientic analysis of childrens testimony. Washington, DC:
American Psychological Association.
Ceci, S. J., Ross, D. F., & Toglia, M. P. (1987a). Suggestibility of childrens memory: Psycholegal issues. Journal of
Experimental Psychology: General, 116, 3849.
Ceci, S. J., Ross, D. F., & Toglia, M. P. (1987b). Age differences in suggestibility: Narrowing the uncertainties. In S. J. Ceci,
M. P. Toglia, & D. F. Ross (Eds.), Childrens eyewitness memory (pp. 7991). New York: Springer-Verlag.
Conducting investigative interviews 821
Craig, R. A., Sheibe, R., Kircher, J., Raskin, D. C., & Dodd, D. (1996). Effects of interviewer questions on childrens
statements of sexual abuse. Unpublished manuscript, University of Utah, Salt Lake City.
Fisher, R. P., & Geiselman, R. E. (1992). Memory-enhancing techniques for investigating interviewing: The cognitive
interview. Springeld, IL: Charles C. Thomas.
Goodman, G. S., & Aman, C. (1990). Childrens use of anatomically detailed dolls to recount an event. Child Development,
61, 18591871.
Goodman, G. S., Aman, C., & Hirschman, J. (1987). Child sexual and physical abuse: Childrens testimony. In S. J. Ceci,
M. P. Toglia, & D. P. Ross (Eds.), Childrens eyewitness memory (pp. 123). New York: Springer-Verlag.
Goodman, G. S., Bottoms, B. L., Schwartz-Kenney, B. M., & Rudy, L. (1991). Childrens testimony about a stressful event:
Improving childrens reports. Journal of Narrative and Life History, 1, 6999.
Goodman, G. S., Rudy, L., Bottoms, B., & Aman, C. (1990). Childrens concerns and memory: Issues of ecological validity
on the study of childrens eyewitness testimony. In R. Fivush & J. Hudson (Eds.), Knowing and remembering in young
children (pp. 249284). New York: Cambridge University Press.
Goodman, G. S., Wilson, M. E., Hazan, C., & Reed, R. S. (April, 1989). Childrens testimony nearly four years after an
event. Paper presented to the Eastern Psychological Association, Boston, MA.
Green, A. H. (1991). Factors contributing to false allegations of child sexual abuse in custody disputes. In M. Robin (Ed.),
Assessing child maltreatment reports: The problems of false allegations (pp. 177190). New York: Haworth Press.
Hughes, M., & Grieve, R. (1980). On asking children bizarre questions. In M. Donaldson, R. Grieve, & C. Pratt (Eds.), Early
childhood development and education (pp. 104114). Oxford, England: Blackwell.
Hershkowitz, I., Lamb, M. E., Sternberg, K. J., & Esplin, P. W. (1997). The relationships among interviewer utterance type,
CBCA scores, and the richness of childrens responses. Legal and Criminological Psychology, 2, 169176.
Jones, D. D. H. (1992). Interviewing the sexually abused child. Oxford, UK: Gaskell.
Jones, D. P. H., & Krugman, R. D. (1986). Can a 3-year-old child bear witness to her sexual assault and attempted murder?
Child Abuse & Neglect, 10, 253258.
Lamb, M. E. (1994). The investigation of child sexual abuse: An interdisciplinary consensus statement. Child Abuse &
Neglect, 18, 10211028.
Lamb, M. E., Hershkowitz, I., Sternberg, K. J., Boat, B., & Everson, M. D. (1996). Investigative interviews of alleged sexual
abuse victims with and without anatomical dolls. Child Abuse & Neglect, 20, 12391247.
Lamb, M. E., Hershkowitz, I., Sternberg, K. J., Esplin, P. W., Hovav, M., Manor, T., & Yudilevitch, L. (1996). Effects of
investigative style on Israeli childrens responses. International Journal of Behavioral Development, 19, 627637.
Lamb, M. E., Sternberg, K. J., & Esplin, P. W. (1994). Factors inuencing the reliability and validity of the statements made
by young victims of sexual maltreatment. Journal of Applied Developmental Psychology, 15, 255280.
Lamb, M. E., Sternberg, K. J., & Esplin, P. W. (1995). Making children into competent witnesses: Reactions to the amicus
brief In Re Michaels. Psychology, Public Policy, and the Law, 1, 438449.
Lamb, M. E., Sternberg, K. J., Esplin, P. W., Hershkowitz, I., Orbach, Y., & Hovav, M. (1997). Criterion-based content
analysis: A eld validation study. Child Abuse & Neglect, 21, 255264.
Memorandum of Good Practice (1992). London, England: Her Majestys Stationery Ofce.
Myers, J. E. B. (1994). The backlash. Thousand Oaks, CA: Sage Publications.
Ornstein, P. A., Larus, D. M., & Clubb, P. A. (1991). Understanding childrens testimony: Implications of the research on
childrens memory. In R. Vasta (Ed.), Annals of child development (vol. 8, pp. 147176). London: Jessica Kingsley.
Paris, S. G. (1988). Motivated remembering. In F. E. Weinert & M. Perlmutter (Eds.), Memory development: Universal
changes and individual differences (pp. 221242). Hillsdale, NJ: Erlbaum.
Perry, N. W., & Wrightsman, L. S. (1991). The child witness. Newbury Park, CA: Sage Publications.
Poole, D. A., & Lamb, M. E. (1998). Investigative interviews of children: A guide for helping professionals. Washington,
DC: American Psychological Association.
Raskin, D. C., & Esplin, P. W. (1991a). Statement validity assessments: Interview procedures and content analyses of
childrens statements of sexual abuse. Behavioral Assessment, 13, 265291.
Raskin, D. C., & Esplin, P. W. (1991b). Assessment of childrens statements of sexual abuse. In J. Doris (Ed.), The
suggestibility of childrens recollections (pp. 153164). Washington, DC: American Psychological Association.
Raskin, D., & Yuille, J. (1989). Problems of evaluating interviews of children in sexual abuse cases. In S. J. Ceci, M. P.
Toglia, & D. F. Ross (Eds.), Perspectives on childrens testimony (pp. 184207). New York: Springer-Verlag.
Saywitz, K. J. (1987). Childrens testimony: Age-related patterns of memory errors. In S. J. Ceci, M. P. Toglia, & D. F. Ross
(Eds.), Childrens eyewitness memory (pp. 3652). New York: Springer-Verlag.
Saywitz, K. J., Goodman, G. S., Nicholas, E., & Moan, S. F. (1991). Childrens memories of physical examination involving
genital touch: Implications for reports of child sexual abuse. Journal of Consulting and Clinical Psychology, 59,
682691.
Spencer, J., & Flin, R. (1990). The evidence of children. London, England: Blackstone Press.
Sternberg, K. J., Lamb, M. E., Hershkowitz, I., Esplin, P. W., Redlich, A., & Sunshine, N. (1996). The relationship between
investigative utterance types and the informativeness of child witnesses. Journal of Applied Developmental Psychology,
17, 439451.
Sternberg, K. J., Lamb, M. E., Hershkowitz, I., Yudilevitch, L., Orbach, Y., Esplin, P. W., & Hovav, M. (1997). Effects of
introductory style on childrens abilities to describe experiences of sexual abuse. Child Abuse & Neglect, 3, 255264.
Undeutsch, U. (1982). Statement reality analysis. In A. Trankell (Ed.), Reconstructing the past: The role of psychologists
in criminal trials (pp. 2756). Stockholm: Norstedt & Sons.
Undeutsch, U. (1989). The development of statement reality analysis. In J. C. Yuille (Ed.), Credibility assessment (pp.
101120). Dordrecht, The Netherlands: Kluwer.
822 M. E. Lamb, K. J. Sternberg, and P. W. Esplin
Walker, A. G. (1993). Questioning young children in court: A linguistic case study. Law and Human Behavior, 17, 5981.
Walker, N., & Hunt, J. S. (in press). Interviewing child victim-witnesses: How you ask is what you get. In C. P. Thompson,
D. Herrmann, J. D. Read, D. Bruce, D. Payne, & M. P. Toglia (Eds.), Eyewitness memory: Theoretical and applied
perspectives. Mahwah, NJ: Erlbaum.
Westcott, H., Davies, G., & Horan, N. (1988, March). Evaluating investigative interviews for suspected child sexual abuse
carried out by English police ofcers. Paper presented to the American Psychology-Law Society Conference, Redondo
Beach, CA.
RE

SUME

Les enfants sont souvent les seules sources de renseignements lorsquon soupconne la maltraitance. Les recherches nous
demontrent que les enfants sont fort competents lorsquil sagit de fournir des renseignements. Cependant, la qualite et la
quantite de ces informations sont largement inuencees par la facon dont les enfants sont interviewes. Dans cet article, nous
passons en revue brie`vement les facteurs qui inuent sur la competence des enfants et nous decrivons les facons dont les
enqueteurs qui me`nent des entrevues peuvent assurer une plus grande qualite des renseignements quils obtiennent des
temoins et des victimes.
RESUMEN
A menudo los ninos son la unica fuente de informacion disponible en relacion a las posibles experiencias de abuso. Las
investigaciones han demostrado que los ninos pueden, en efecto, ser informantes extremadame competentes, a pesar de que
la calidad y la cantidad de la informacion que ellos ofrecen est muy inuenciada por las formas en que son entrevistados.
En este articulo revisamos brevemente los factores que inuyen en la competencia de los ninos y describimos formas en
las que los entrevistadores pueden maximizar la calidad de la informacion que obtienen de supuestos testigos y v ctimas.
Conducting investigative interviews 823